Why are we afraid of robots? Containers and the Uncanny Valley
Looking at the long-described phenomenon from a different (Freudian) perspective.
File this post in the folder of things that I was going to tack on as an addenda to my last post on ‘Why we’re afraid of robots’ but, as is often my way, that post got a little too long as it was, and then this idea started turning into a Thing in its own right, so…
This idea is a little more… well, let’s say ‘involved’. We need to drop deeper into some psychoanalytic theory and look at culture through this lens to set it up. So even if this is a little more specific than I would normally like to post up here, I think that this one is worth it, because of the different paths it invites us down.
Projections and containers revisited
So, just to go back to Freud for a bit — and who’s ever had enough of Freud? like, no one ever, I’m sure — we can add some more complexity to the concept of projections, which I introduced in a previous post on the question of why we are afraid of robots. Projections are central to so many of our fears and the way we think about the world more generally that they’re bound to keep popping up.
Projections are a defence mechanism, but they are also part of normal human psychological functioning. There are plenty of examples of people using projections and containers pathologically, or for negative ends (e.g. racism). But we can also use projections in everyday life, as a means (unconsciously) of communicating and interacting with others in many positive ways. We subtly, unconsciously, ask others to contain or hold our good and bad parts of ourselves, and pass them back to us, when we’re better able to take care of them ourselves.
Good parts are projected elsewhere when we feel like that part of ourselves might be under threat. Or we identify with the bit of ourselves contained by others, or groups, or even contained by society itself. For many psychoanalysts who work on groups and collective phenomena, containment is the basis of a collective identity. We identify with the good and bad things about a group, things that we imagine we’ve put there from ourselves.
Bad parts can get projected out in numerous ways, too. We might project our own feelings of jealousy onto others, like our partners, and accuse them of having the sort of cheating heart that we’d rather not admit is actually our own. Or we might project our hateful impulses on another race, accusing them of hating us and trying to kill us, while telling ourselves that they ‘hate us because of our freedom’. Sometimes we project these bad things elsewhere and try to banish them forever, or destroy the container holding our bad parts of ourselves altogether, like that poor scapegoat.
Sometimes, though, we project the bad parts outwards and put it in a container, hoping that these bad parts can be sanitised or cleansed. This is probably a more everyday sort of thing, and certainly nothing that we’d want to pathologise (like racism and genocidal impulses). This is the kind of work that parents do for their toddlers. The child’s temper tantrum is frustration put out into the world. If the parent can take on these negative feelings and show their child that frustration is ok, neutering (or ‘nerfing’) the bad feelings, it shows the child that these things aren’t so bad after all. The parent contains the negative projections, the anxiety that the child can’t tolerate being inside, and neutralises their power. The child is better able to deal with the anxiety, and learns important coping mechanisms.
Robots as an (unsuitable) container
Robots, like most monsters, as I explained in my previous posts (here and here), are containers. They are the creation of our (human) projections. They are, very essentially, us, or at least versions of us. They look like us, not just physically, but they contain some of our worst impulses. Robots can be genocidal, bent on world domination. Robots are our rational selves, divorced from affection and love and the human connection that our modern world makes us afraid of losing. (Robots can also represent some positive aspects of ourselves, like maybe our belief that we can invent new things to solve our problems. I might discuss this one day in the future, but it’s not very ‘monstrous’ so I’m not rushing to cover these things. But file that under the ‘messiahs’ part of the machines for now.)
Sometimes robots look so much like us that we can’t tell us apart. That’s a key plot element of a lot of our robot-based sci-fi. And I mean a lot. A highly selective list, chronologically, might include: R.U.R. (1920), Metropolis (1927), The Stepford Wives (1974 & 2004), Westworld (1973 & 2016 -22), The Terminator series in all its manifestations (1984 - present), A.I Artificial Intelligence (2001), Battlestar Galactica (2004 - 2009), Ex Machina (2014). And no, I didn’t forget Do Androids Dream of Electric Sheep (1968) and the cinematic versions, Blade Runner (1982 & 2017).
The fear that we cannot tell the difference between man and machine is an existential fear, not just because we cannot identify what it is that is ‘human’ and it makes us question, on a fundamental level, what we are, but also because we are unsure who to trust with our projections. An unsuitable container, one that cannot receive, process, or return projections adequately, can have dire consequences for the integrity and conception of the self.
In Do Androids Dream?, Deckard very explicitly explains that it is this, at least in part, that is responsible for his hatred of andys, and of the ‘electric sheep’:
He thought, too, about his need for a real animal; within him an actual hatred once more manifested itself toward his electric sheep, which he had to tend, had to care about, as if it lived. The tyranny of an object, he thought. It doesn’t know I exist. Like the androids, it had no ability to appreciate the existence of another.
Tying back to the idea, too, that humans are becoming cold, mechanical shells, humanoid robots remind us how close we are to inhumanity ourselves. They show us what we might become as we lose our capacity for emotional engagement with our world and with other humans. Unable to give or take or process projections or have them processed for us, we become isolated, depersonalised, depleted of our humanity, empty of those good parts of the self that enable us to empathise and engage with the world beyond reason.
Taking a dip in the Uncanny Valley
We can tie this to another idea, one you’ve likely already heard about. The theory of the uncanny valley has long maintained that it is the robots that look most human that are regarded as the most frightening. But why? The idea of projection provides us with another answer — which is not, of course, as always, to discount any others. We stand on the precipice of the uncanny valley when robots appear human and we are tempted to engage with them as humans, rather than as machines, or are unsure how to relate to them altogether. And when the robot most closely approximates a human, we feel a special anxiety projecting those parts of ourselves that make us vulnerable to depersonalisation and disintegration; that is, when the limitations of the machine threaten our own self, the fabric of our being.
(This connects us in a small way to Freud’s own notion of the uncanny, Unheimlich: what threatens us is the unthought known, the reflection of self that we cannot accept as the self, that we dare not acknowledge. But I don’t want to push this connection too far because, while we are literally using the same word for both phenomena, they are very different things.)
I suspect — and at this point, it’s just a suspicion; I need to do more research — that one reason for the uncanny valley, from nameless discomfort to paralysing terror, is that we don’t know what to do with these creatures, these robots. They look like us, they invite our projections. But sticking things in there isn’t like projecting into your mother, or a friend, or a charismatic leader, or a group. These mechanical containers are like coffins. Projections go there and aren’t processed. They either go to die, or get lost or crushed amidst the servos and the gears, or — most terrifying of all — like the electricity that brings Frankenstein’s monster to life, our projections can animate a mechanical horror, a simulacrum of a man, making alive our worst impulses and fears.
We see this story again and again in representations of robots. And like the monster in Mary Shelley’s gothic horror, there is a warning here about the consequences of reason. So many of our monsters since the nineteenth century – Frankenstein, Mr. Hyde, Nazism, zombies and robots – are the result of our technology and our science without our humanity there to mediate or moderate us. H.A.L. 9000, The Terminator, the Borg are ruthless in their efficiency, monsters made all the more destructive and potent by the fact that they are guided by a single principle – not an irrational violence, but a violence that is completely and utterly rational, based in a calculated, indisputable logic, a fanatical dedication not to myth (as with the savage or the religious extremist) but to a technological ideal. to programming. These monstrous machines have no time for our Romantic notions about the sanctity and specialness of human life. Patrick Stewart can quote Shakespeare from the bridge of the Enterprise and convince us all that we’re just wonderful, despite all of our faults, but the Borg are never going to fall for it (if you know, you know). Such monsters are the embodiment of our bad selves, the dreaded consequences of our reason, our science and our technology.
Next thoughts and further reading
This has led me back down some old psychoanalytic paths, roads I once knew so well and now are overgrown with weeds of forgetting. I’ve been looking at some of these ideas in more detail: André Green and the idea of ‘dead mother’, Wilfred Bion and the ‘nameless dread’, Robert M. Young, Melanie Klein, D. W. Winnicott… If you’re so inclined and want to read more on these ideas in their psychoanalytic contexts, that’s where you should start. I’ve already reached into the back of my bookshelf (yes, books in my tiny little office are stacked three deep) and started to flip through some of these old pages, with scribbles all over the margins of the browning pages written by someone that is only vaguely familiar to me today; he seems silly a lot of the time, but he might just have been on to something.
These psychoanalytic ideas aren’t sexy any more. Like, at all. But they’re still useful. They tell a really powerful story in some very compelling ways.